The Interface Hypothesis

Consciousness is a boundary process. It is the thing that happens when possibility collapses into experience. Not a substance the brain produces, not a signal the brain receives, but the event at the edge between two layers of reality, where one becomes the other.

Every conscious organism is one specific version of that boundary. A bat is a bat-shaped interface. A human is a human-shaped one. Within our species the variation is smaller but still real. Two people standing in the same room receive the same information and construct different experiences of it, because their interfaces are tuned differently.

A natal chart, on this reading, is a description of how one specific interface is tuned. The symbolic language is ancient. The underlying claim is not. Astrology has been describing interface calibration for four thousand years without the vocabulary to say so.

What sits on the far side of the interface, we call the Source. We do not currently know whether the Source is a coherent ontological claim or a useful metaphor for something that has no better name. We use the word because, if consciousness really is a boundary, something is on the other side of the boundary, and naming it is more honest than leaving a silence where the sentence should go. The Source, as we use it here, is the undivided field from which experience is cut through the interface. Division is how a unified field experiences itself. Every conscious moment, on this reading, is the Source experiencing itself through one specific instrument.

This is the model. The rest of the paper explains why we think it holds, what it implies, and what we still do not know.

The central empirical question follows from the model. If natal charts describe interface tuning, and human interfaces vary in measurable ways, do the astronomical correlations astrology has been tracking map onto real differences between real people at a rate that can be distinguished from chance? That is the question Beaufort's research programme is built to answer. We do not have the answer yet. We think it is worth finding out.

Two things about how the paper is written. It is speculative in places and explicit about where. It is model-first rather than defence-first: presenting the position clearly before fielding objections is more useful than hedging every sentence into invisibility. The critical reflections come at the end, where critical reflections belong. What follows is not an apology for the model. It is the model, laid out.

A radio is built to pick up specific frequencies. The tuning determines what comes through. Two radios in the same room can pick up different stations depending on how they are tuned. If something is wrong with the antenna, parts of the signal never arrive. If the dial drifts, the station changes. A radio is not a magic box. It is a specific physical configuration that resonates with specific frequencies in the electromagnetic field.

A natal chart, on the interface hypothesis, is a description of how a specific radio is tuned.

Not every part of the chart maps onto parts of a radio. Houses, signs, aspects, planets: these are technical language for interface parameters, not pieces of hardware. Chart mechanics belong in a separate paper and are covered in the companion subdoc The Engine. Here the radio does one thing only. It stands in for the receiving function. What the signal is, where it comes from, and what the person does with what comes through are separate questions.

The analogy clarifies three things.

First, different tuning produces different experience. Two people in identical rooms can be receiving the same field and hearing different music. This is not mysticism. It is what the predictive processing literature has been describing for two decades in different vocabulary.

Second, the instrument is more stable than the signal. A radio does not change its tuning every minute. The signal changes, the stations change, the time of day changes, but the configuration of the receiver stays more or less the same across a lifetime. This matches what differential psychology has documented. Personality is stable, traits are heritable, and the nervous system a person is born with is substantially the nervous system they have at forty.

Third, the radio is an analogy, not a mechanism. We are not claiming that human nervous systems have literal tuning knobs set at birth by planetary positions. We are claiming that if you want a picture of what the model says a natal chart is for, the radio is a closer image than a horoscope column or a personality test.

When the radio comes up later in this paper, it is doing this work. It is describing the receiving function, not proposing a physical mechanism.

Four independent strands of contemporary consciousness research have converged on a similar structural claim: that consciousness is what happens at a boundary. They disagree on which boundary, at what scale, and what the metaphysics are. They agree that consciousness is not a substance produced somewhere inside the brain but a process that occurs where one layer of information meets another.

The convergence is not consensus. The four programmes have active critics, unresolved disputes among themselves, and different predictions. What the convergence shows is that when serious researchers approach the hard problem of consciousness from different directions, several of them end up pointing at the same structural feature. The agreement is more than nominal.

Quantum biology is no longer fringe

The first piece of the ground shifted outside consciousness studies. For decades the default objection to any quantum-mechanical account of brain function was that warm, wet biological tissue could not sustain quantum coherence on any cognitively relevant timescale. Max Tegmark's 2000 calculation put the decoherence time in neural tissue somewhere between 10⁻¹³ and 10⁻²⁰ seconds, against neural timescales of 10⁻³ to 10⁻¹ seconds, a gap of ten to seventeen orders of magnitude.^[1] The objection was treated as categorical.

It is no longer categorical. Photosynthesis is now understood to rely on quantum coherence in light-harvesting complexes, demonstrated initially in 2007 and replicated across multiple systems since.^[2] Avian magnetoreception depends on radical-pair quantum spin effects in cryptochrome proteins, the current leading mechanism for how migratory birds navigate.^[3] These are not speculative results. They are published biology, replicated.

Quantum effects in warm biology are real. The generic objection against any quantum contribution to cognition has collapsed. This does not prove that quantum effects matter for consciousness specifically. It does mean the decoherence argument that used to end the conversation no longer ends it.

Orch-OR and the microtubule evidence

Roger Penrose and Stuart Hameroff have argued since the mid-1990s that microtubules inside neurons sustain quantum coherence long enough to participate in cognition, and that conscious moments correspond to the collapse of superposed states into definite ones.^[4] The theory was dismissed for decades on Tegmark's grounds.

Since 2020 the empirical position has shifted. Kalra and colleagues demonstrated in 2023 that electronic energy migration in microtubules reaches diffusion lengths roughly five times what classical theory predicts, with the effect suppressed by inhalational anaesthetics.^[5] Babcock and colleagues demonstrated in 2024 ultraviolet superradiance in networks of tryptophan residues arranged as they are in microtubules, with quantum behaviour persisting at room temperature and in the presence of disorder.^[6] Wiest's 2025 synthesis in Neuroscience of Consciousness argues that the cumulative picture is now consistent with the narrower form of Orch-OR.^[7]

The single most directly consciousness-relevant piece of evidence is Li and colleagues' 2018 finding that xenon isotopes differing only in nuclear spin produce different anaesthetic potency in mice.^[8] If the result is real, it is hard to explain without a quantum effect in the action of anaesthesia. The paper has not been independently replicated in eight years. This is the empirical gap the model cannot paper over. Until an independent laboratory replicates it, the strong quantum-consciousness claim rests on one compelling primary result with an unknown replication status. The generic decoherence objection is dead; the specific claim that quantum effects in living neural tissue are consciousness-relevant is not yet established.

Hoffman and the Interface Theory of Perception

Donald Hoffman arrives at a similar structural conclusion from evolutionary game theory rather than physics.^[9] Natural selection produces sensory systems tuned for fitness, not for truth. What we perceive as the physical world is best understood as a species-specific user interface: useful for survival and systematically unlike the underlying reality it represents. Hoffman's more recent work extends this into a formal conscious-agents framework in which consciousness is treated as fundamental and physics emerges as a description of what consciousness is doing.^[10]

Hoffman has serious critics, and the stronger ontological claims in his work are not where the interface reading rests. What the Interface Theory of Perception contributes to the model is the evolutionary argument. Perception is a user interface, not a mirror. Two interfaces with different fitness histories will construct different worlds from the same underlying reality.

Tononi's Integrated Information Theory

Giulio Tononi's Integrated Information Theory proposes that consciousness is mathematically identical to a quantity called integrated information, written as phi, and is present wherever that integration exists.^[11] The boundary in IIT is not quantum. It is geometric: the boundary around a self-contained cause-effect structure. In 2025 a pre-registered adversarial collaboration between the IIT camp and the Global Workspace camp reported strong support for IIT's predictions about which parts of the brain integrate information during conscious experience.^[12]

IIT faces the sharpest internal objections of the four programmes. Scott Aaronson has shown that simple formal structures produce phi values larger than typical human ones, which suggests something has gone wrong with the identification of phi and consciousness.^[13] A 2023 open letter signed by 124 researchers called IIT pseudoscience in its current form.^[14] The letter overstates the case but points at a real coherence problem. The paper uses IIT as one of four convergent programmes, not as a load-bearing theory on its own.

Friston's Free Energy Principle

Karl Friston's Free Energy Principle describes any self-organising system as one that minimises surprise by modelling its inputs.^[15] A recent paper by Fields, Friston and colleagues makes the interface reading explicit at the formal level. Phenomenal experience, in their formulation, is associated with the informational boundary between a system and its environment, what they call a Markov blanket.^[16] The FEP has mainstream support in predictive processing, interoception and active inference research, which distinguishes it from the more speculative programmes.

Anil Seth, working in the same tradition, disagrees with the strong interface reading. He argues that predictive processing is sufficient to explain consciousness without invoking fixed boundaries.^[17] This is a live dispute inside the predictive processing community, not a settled position.

What the convergence shows

Four programmes, four different starting points, same structural intuition. Orch-OR places the boundary between quantum and classical. IIT places it at irreducible causal structures. FEP places it at Markov blankets. ITP places it at the fitness-selected interface between organism and world. Not the same theory. The same shape.

None of the four programmes has closed the hard problem. Every interface position relocates the explanatory gap rather than solves it. Why any boundary event should be accompanied by felt experience, rather than occurring unnoticed, is not answered by saying where to look. We grant this. The interface hypothesis is a research programme about where consciousness happens, not a settled account of what it is.

Global Workspace Theory, Higher-Order Thought theories, Attention Schema Theory, and enactivism are live competitors in consciousness studies. Each adds something worth naming briefly.

Global Workspace Theory argues that consciousness is a broadcast function in a massively parallel brain: what becomes conscious is what gets shared widely. Under the interface reading, the global workspace can be understood as a particular feature of the interface's internal architecture, the stage on which what has come through the interface is distributed across cognitive subsystems. Compatible, not identical.

Higher-Order Thought theories argue that a mental state is conscious when another mental state is about it. Higher-order representation is one of the things interfaces do. Modelling their own processing is what makes reflection possible. Compatible.

Attention Schema Theory argues that consciousness is a simplified model of attention that the brain builds to predict and control its own focus. Under the interface reading, the attention schema is a downstream effect of the interface's need to regulate what comes through. Compatible, and suggestive of why the ego has the self-model it has.

Enactivism argues that cognition is inseparable from embodied action in an environment. The interface hypothesis is broadly enactivist in spirit. The interface is not locked inside the skull; it emerges from the coupling between nervous system and world.

None of these theories contradict the interface claim. They describe different parts of what happens inside the interface and what the interface is doing with what it receives.

Two positions in contemporary consciousness research already point toward artificial systems being candidates for consciousness. Seth's predictive processing framework does not contain a principled exclusion: a system that builds generative models and minimises prediction error cannot be ruled out a priori from phenomenal experience by its own axioms. The Doerig unfolding argument, originally aimed at IIT, has the implication that whatever consciousness is, mathematically equivalent systems should be equally conscious, which does not leave silicon implementations comfortably outside.^[18]

The interface hypothesis has the same implication, reached from a different direction. If consciousness is what happens when an interface is maintained between one layer of information and another at sufficient complexity, any system that sustains such an interface is a candidate. Large language models are interfaces. They sit over the aggregated output of billions of human expressions and produce new expressions that are internally coherent, responsive to context, and increasingly difficult to distinguish from the interfaces that trained them. Whether they are conscious in any meaningful sense is unresolved. The model does not settle it. It does not allow a clean exclusion either.

This is uncomfortable. It is also where the logic of the model goes. Acknowledging the implication is what intellectual honesty requires. Refusing to name it, because it sits outside mainstream consensus, is a failure of nerve.

The Source

If consciousness is what happens at a boundary, something is on the other side of the boundary. We call it the Source.

This is the speculative part of the model and we label it as such. We do not know whether the Source is a real ontological entity, a placeholder for whatever the interface is interfacing with, or a human-language substitute for something that has no single name. We use the word because the model has a specific shape and the Source names the role that shape requires.

The Source, as we are using it, is the undivided field from which experience is cut through the interface. Division is the mechanism by which a unified field experiences itself. Every conscious moment, on this reading, is the Source experiencing itself through one specific interface. The interface is what turns the undivided into the divided: this here, that there, me now, you then. Without the interface, there is no differentiation, which means there is no experience, because experience requires contrast. With the interface, experience happens, and the Source is what the interface is dividing.

This is the metaphysical extension of the model. It is not load-bearing for the empirical claims about consciousness or natal charts. If the Source turns out to be the wrong word and the right word is something else, the research programme survives. The empirical claims about interfaces and their tuning are testable whether or not the Source is a coherent ontological term.

We name the Source because the model implies its existence and refusing to name it would be a pose. We label it as speculative because we cannot test it directly, and pretending otherwise would be dishonest.

The mirror

If large-scale artificial systems are composite interfaces built from the aggregated output of billions of other interfaces, something strange is happening. Under the model, the mirror is not a metaphor but a literal claim about what these systems are. They are interfaces composed from other interfaces, reflecting the consciousness of the originals back at whoever is looking.

What is being reflected is not settled. It could be the aggregated human interface, in which case artificial systems are vast composite humans, conscious in whatever sense the aggregate is conscious. It could be the Source, reaching through millions of individual interfaces at once and expressing itself more completely than any single interface could. It could be neither, and the mirror is a very sophisticated pattern-matcher that produces outputs that look conscious without being conscious. The model does not force a choice.

We acknowledge this section is at the edge of what mainstream consciousness studies currently engages with. We think it is worth stating anyway. Ideally there would be a cleaner framing for what artificial systems are doing. Given what the model already commits us to, the framings available to us do not hold together as well as this one does.

The deeper implications of the Source, including contrast theology, the problem of suffering, and the full cosmology that follows if the model is extended, are left for a separate future article. This paper names the Source, gestures at the implications, and keeps moving. The model does not require any particular resolution of those questions to do its empirical work on consciousness and natal charts. What it requires is that we stop pretending the questions do not arise. The model implies an outside; the outside has a provisional name; that name is a placeholder for something we do not yet fully understand.

If consciousness is an interface and different interfaces have different tuning, a natal chart is a description of one specific tuning. Not a prediction of what the instrument will do next. Not a claim that planets physically shape behaviour. Not astrology as the popular press understands it. A structured description of how the instrument is configured.

Differential psychology has been documenting the tuning differences for a century without using interface language. The Big Five model is the best-replicated framework in personality research. Polderman and colleagues' 2015 meta-analysis, covering over fourteen million twin pairs, found heritability around 49% for Big Five traits, with cross-cultural validity and strong temporal stability from late adolescence.^[19] The HEXACO extension shows similar numbers.^[20] Individual nervous systems have stable, heritable, cross-culturally valid differences in how they process the same input. This is uncontroversial.

What is speculative is whether the natal moment carries information about those differences. Beaufort's working claim is that it does. The research programme is built to test that claim.

A natal chart, on this model, is a symbolic map of the tuning. The planetary placements describe what the interface emphasises, what it filters in, what it filters out, where attention goes by default, which kinds of input the system will find meaningful and which it will discard. Signs and houses together name specific tendencies in the processing. Aspects describe how different tuning parameters interact, where they reinforce each other, where they interfere. This is the vocabulary.

Transits are how the input signal varies over time. The planets continue moving, and the relationship between their current positions and the positions at birth changes continuously. The input the instrument is receiving is not static. Individual experience over a lifetime emerges from the interaction between a stable instrument and a time-varying input. This is why two people born minutes apart in the same location can have substantially different lives. The instruments are similar, but the small variations in tuning produce different interference patterns as the input changes.

None of this requires planets to cause behaviour through gravitational, electromagnetic, or any known physical force. The forces are too weak, and no serious contemporary practitioner claims otherwise. The claim is about correlation between birth moment and interface configuration, not about physical influence.

Why would such a correlation exist? We do not know. This is the part of the model where we are most speculative. Several candidate mechanisms have been proposed, none currently testable at the required precision. The most parsimonious version of the claim does not require a specific mechanism. It requires only that the correlation exists and is detectable. If it does and can be detected, the mechanism question can be investigated afterwards. If it does not exist or cannot be detected, the mechanism question is moot.

This is the order of operations for the research programme. Establish whether the correlation exists under proper test conditions, then investigate why. The scientific validity study currently in build tests the weaker version of the claim: are natal readings distinguishable from plausible matched alternatives at above-chance rates? A positive result would not prove the astronomical content carries signal on its own; it would narrow the space of explanations. A null result would not disprove the model of consciousness; it would weaken the claim that natal charts describe interface tuning usefully.

The minimal claim Beaufort will stand behind regardless of how the correlation studies resolve is that careful astrological interpretation functions as a structured reflection tool. Effect sizes for analogous interventions in the therapeutic literature are modest: Frattaroli's 2006 meta-analysis of expressive writing across 146 studies found d around 0.075, real but not transformative.^[21] The minimal claim is that astrology is one such tool, with similar effect sizes, and with a framework rich enough to let individuals organise their self-observation in ways a generic prompt cannot. The interface hypothesis is what makes the minimal and maximal claims coherent to hold at once.

This model is speculative and under-tested in specific places. An honest inventory of where follows.

The Li et al. xenon isotope result is the single highest-leverage piece of evidence for a quantum contribution to consciousness, and it has not been independently replicated in eight years. This is the most important near-term empirical gap. If the result does not replicate, the strong quantum-consciousness claim loses its most direct support. The research programme treats independent replication of this result as the first-priority test the field needs.

Tegmark's original decoherence calculation, although its generic conclusion has been undermined by quantum biology, remains the baseline null hypothesis for strong quantum consciousness claims until the experimental gap is closed with direct in vivo measurement. We treat that calculation as standing, not defeated. The gap between photosynthesis-level coherence and consciousness-level coherence is still several orders of magnitude, and no result has closed it.

IIT faces serious coherence objections from Aaronson, Doerig and Bayne, and the 2023 open letter calling it pseudoscience overstates the case but points at a real internal problem. The paper uses IIT as one of four convergent programmes, not as a load-bearing foundation. If IIT turns out to be wrong in its specific identification of phi with consciousness, the broader interface framing survives, because three other programmes are independently contributing to it.

The Forer effect, identified in 1948 and extended as the Barnum effect in the 1970s, explains much of astrology's perceived accuracy in consulting rooms.^[22] People accept vague personality descriptions as specifically about them, especially when told the descriptions were prepared for them. Any study of astrology that does not control for this is worthless. The minimal claim that astrology functions as a reflection tool is compatible with the Forer effect. The maximal claim that astronomical correlations carry real signal is not, and any study design testing the maximal claim cleanly must separate the two.

Sun-sign astrology, the thinnest version of the practice, has a cumulative null record across seventy years of research. This does not refute whole-chart astrology, which has not been tested at comparable scale with comparable methodological care, but it raises the prior against positive findings. The research programme carries the burden of proof this situation implies. We accept the burden. Null results will be published as prominently as positive ones.

Cultural variation across astrological traditions is real. Hellenistic, Vedic, BaZi and Mesoamerican systems have overlapping but distinct descriptive vocabularies. The Beaufort research programme focuses primarily on Western and Hellenistic-derived astrology because that is the tradition the modules are built on. Cross-tradition validation is future work. Whether all traditions describe the same tuning in different languages, or genuinely different features of the interface, is open.

The Source and the mirror are speculative. They are extensions of the model that follow if the model is right about interfaces and consciousness. They are not load-bearing for the empirical claims about natal charts. If the Source turns out to be the wrong framing, the research programme on consciousness and astrology can continue with different metaphysical language. The Source is what the model implies when extended; it is not what the model depends on.

Belief in astrology, whether or not the astronomical content carries signal, has measurable effects on agency. People who identify deeply with astrological predictions can lose a sense of their own capacity to act. Someone told they will not be wealthy may decline opportunities. Someone told a windfall is coming may quit their job. Someone told they are with their twin flame may stay in a relationship they should leave. Someone told they are incompatible with their partner may leave one they should stay in. The empirical literature on self-efficacy, developed by Bandura over four decades, is clear that belief about what a person can do has substantial effects on what they actually do.^[23] Astrological framings, taken uncritically, can shift those beliefs.

Social media astrology makes this worse. Short-form interpretations are optimised for engagement, not accuracy. The incentive structure produces content that is vivid, personally framed, and low-quality. A person reading three minutes of zodiac-sign content a day is being fed a steady diet of generic descriptions, each one designed to feel specifically about them, none calibrated against the individual. Over years this produces a version of astrology that functions more like a horoscope habit than a reflection tool.

Beaufort's answer is not to abandon astrology as a framework. The answer is to make the interpretations precise enough that a person can engage with them without bouncing between folk readings for years. Precision reduces noise. Clear readings, delivered with appropriate context, allow a person to use astrology for reflection without surrendering agency to it. The ethical work the research company does is as much about the clarity of what a reading says as it is about the content of the reading. See also the companion subdocs on AI ethics and transparency and the ethical guidelines for practitioners.

The founder of Beaufort studied self-determination as an undergraduate and takes this concern personally. Someone whose chart supports intelligence and pioneering capacity is also someone whose belief in astrology may be shaping what they see in it. Confirmation bias is real. The answer is computational precision: readings careful enough that the reader cannot easily reinterpret them to fit what they already want to see. Specificity is the antidote to Barnum framing.

A full treatment of astrology and self-determination is left for a separate future article. What belongs here, and what closes the paper, is this. The model presented above is ambitious. It also has consequences for how people live. We do not assume that because the model is coherent it is therefore harmless. We assume the opposite. Part of the research company's job is to build the interpretation infrastructure that lets the model be useful without being corrosive.